Small UAV Design

by Red20RC | March 14, 2014 | (14) Posted in Just Fun

After finally getting a chance to do some more test flying (and filming) of the foam board UAV, I sat down and thought about the difficulties in successful small UAV design – and why the foam board UAV may NOT be the answer…

Let’s start by watching the foam board UAV flight video:

My apologies for the quality of some of the footage – the GoPro was filming at 720/60, which isn’t the best for distance work. You can understand though why I have been so far reluctant to put my APM controller into this model!

So is full-size UAV design the best pattern for a small UAV?

There is actually no real reason why a full-size UAV design would NOT work on a model. Outside of the world of UAVs and Drones, there are thousands of scale flying models out there that fly perfectly well every weekend. So why did my Foam Board UAV design give me so many problems?

I think the first and most important reason is that the Foam Board UAV is NOT a scale model of any full-size drone. Yes, it took inspiration from the “Predator” and “Reaper” drones but at no stage did I look at a 3-view (like the one above) and take measurements that were then translated to the model design. Let’s take a look then at some of the problems I faced and see if we can learn anything from the real thing…

1. Difficulties in hand-launching

A pusher motor is always going to give you problems when you hand-launch a model. Of course the full-size has long, retractable undercarriage. If I had fitted undercarriage to the model then I wouldn’t have experienced a problem.

2. Vertical take-off performance

Again, not really a full-size comparison issue. The model was badly setup initially and very overpowered. Both the Predator and the Reaper are pusher-prop driven aircraft although neither have more power than they need. UAVs of this kind are built for economy and endurance, not 3D aerobatic performance!

3. Difficult to make smooth, coordinated turns

This is possibly my first obvious design flaw. in the 3-view above you can see that the wing is mounted at around 50% of the fuselage length. This allows for plenty of room at the front of the aircraft to store heavy imaging equipment away from the engine systems. This reduces the possibility of interference but also make it easier to balance the airframe.

Compare this then to my UAV design and you can see that I have taken it to the extreme. The wing sits at around 66-70% of the fuselage length, leaving little or no distance between wing and tail surfaces. The outcome of this is that the tail moment-arm is very small and the yaw effect of the stabilisers is minimal. Moving the wing forward, even by a few inches, would greatly improve the turn performance.

4. Sudden unexpected maneuvers and wing flex

These two are linked and don’t really reflect full-size design. The truth of the matter is that foam bends. As soon as airspeed increases and the nose pitches down a long flexible wing will bend down at the tips. On a design such as this with very long and tapered wings this anhedral flex is pronounced and the effect on flight performance profound. The solution here is to simply increase the strength of the wing and/or reduce the wingspan.

5. Poor glide performance

Even with the large equipment “bubble” in the nose, the Predator is still a very sleek aircraft with a low drag coefficient. The Foam Board UAV however was designed with simplicity in mind and those curved cheeks hide a large flat plate that was used to mount the camera. Put a large flat plate on the front of any small model aircraft and you are going to increase drag dramatically – and as we all know, drag is the arch enemy of glide performance!

So what IS the best small UAV design?

There is still a lot of merit in the Foam Board UAV design but I keep asking myself if it will ever truly deliver the results I had looked for in the beginning. When searching for a successful small UAV design (fixed wing) there are a few key points that it simply must address – in my humble opinion:

  1. Size – Obviously larger models tend to perform better than small ones but we always need to consider how we are going to transport the model. The wing on my Foam Board UAV design was nearly 2 meters and in one piece, which was not easy to move!
  2. Power - Whilst we don’t want amp sucking 3D power-to-weight ratios, we do want enough power to get up there quickly and out of trouble if required. Flight times are limited by battery capacity and spending the first 2 minutes of any flight climbing to altitude is going to badly effect mission success.
  3. Launching – Small UAVs rarely perform missions within the small radius of the local airfield (or even some short grass for that matter). Taking off and landing on spindly undercarriage is simply not practical so a small UAV design MUST be able to be hand-launched with a near 100% success rate.
  4. Performance – Every control input effects performance and reduces flight efficiency (just ask the DLG guys). A small UAV ideally needs the minimum number of servos, driving the minimum number of control surfaces over the smallest distance. All this shouldn’t effect flight performance though and the UAV needs to be stable (especially if you are filming or mapping), have a gentle stall and a reasonable glide. What it doesn’t need is the ability to snap roll, flat spin, hover or fly sustained inverted!
  5. Space – UAVs need to carry a lot more gear than your average RC model so extended internal space for efficient layout of electronics is essential.
  6. Durability – Operating in often harsh conditions means a small UAV needs to be able to take the odd knock without falling to pieces.

What’s out there already?

Do a quick image search on Google for “Small UAV” or “Model UAV” and see what you get. A few designs are immediately obvious…

The most obvious thing we see is that all small UAVs have:

  1. NO undercarriage
  2. Pusher motor configuration

A lot of the time UAV and FPV go hand-in-hand and it is the necessity of a clear forward field of view that has driven the pusher revolution.

Of the more traditional designs present we are seeing a lot of “pod and boom” UAV design. This can either be a single boom with a high wing (as in the original Skywalker), or a twin boom (as in the Aersonde). My only concern here is durability. Unless you are going to invest in expensive build techniques and materials I worry that a flex free, durable boom would be difficult in a small UAV.

Where these designs really wing though is the ability to break them down for transport. Removable two-piece wings are a great advantage in small UAV design. It is even better if the often flimsy tail feathers can be broken down for storage as well.

So now we turn to the wings. I have to admit that the more I think about it, the more I like flying wings for UAV design.

…and here’s why:

  • Wings can provide a large surface area within a relatively small wingspan.
  • When setup correctly they are smooth and efficient aircraft with good glide performance.
  • They are relatively easy to launch once trimmed and can be solo launched as well.
  • Some wings have a lot of internal space in their design, making it easy to space out electronics. This is espescially true of designs such as the Skywalker X8 or “snub nose” wings like the Zephyr II.
  • Wings are inherently strong being one piece formats with no thin or flimsy booms.
  • With only two moving surfaces and two servos, wings can be economical on control inputs and power drain.
  • They look cool :)

Conclusion

For now I have decided not to pursue the foam board UAV design any further, although the plans remain available for those that want to have a play. I will be focusing my attentions on pushing the envelope of what is possible with flying wings – particularly those that can be scratch built from cheap materials.

I hope you have found this rather long article interesting and maybe even thought provoking. I look forward to hearing your views in the comments!

You can learn more about the UAVs I design and build plus lots of other great content at www.red20rc.org

COMMENTS

pintokitkat on March 14, 2014
I wonder what separates a UAV from any other design of plane that we all chuck in the air. After all, none have a pilot on-board, so to that extent, they are all UAVs.
Log In to reply
Red20RC on March 14, 2014
That is an awesome comment and something I didn't touch on in the article.
I'm sure a lot of people will have different opinions and definitions. My own feeling is that a UAV/Drone is a pilotless vehicle with a defined purpose or mission.
What does entertain me is the thought that 5 or 10 years ago if you put a camera in a plane it was "a RC plane with a camera in it". Now of course the media has labelled it a "drone" and look at the trouble that has caused!
Log In to reply
jollyboy1 on March 15, 2014
An RC aircraft with a camera is still an RC aircraft.

Problems arise when people put in autopilots and FPV systems and fly above the 400ft altitude ceiling and beyond line of sight. There are multiple videos on youtube of people flying above a 1000ft upto 60km away from the base station. The sky isnt as big as we think it is.

http://en.wikipedia.org/wiki/Mid-air_collision

Log In to reply
Red20RC on March 16, 2014
Agreed (and I'm definitely not getting into the "drone" debate here!).

There is different legal aspects depending on which country you are in. Of course, we are all watching the debacle in the States with the FAA with great interest.
Down here in Australia things are more relaxed. For example, if you are flying an RC aircraft (non-commercially) away from a "populous" area, you are allowed to go over 400ft AGL as long as you have LOS (as defined in CASA 101.400). At our club field I think the ceiling is 3000ft.

To fly a "UAV" commercially is slightly different. It is possible (although ridiculously expensive) to get a license from CASA to operate a UAV for commercial purposes. Unfortunately there are only a few training companies that have been appointed by CASA and the nearest to me seems to have taken this approval as a license to print money - i.e. around $10k for a certificate that allows 400ft AGL/LOS operations (known as the "Bronze Wings" in recreational flying here). They are also in the development phase at the moment for a "XLOS" certificate that will allow beyond line of site operations, probably whilst still under a data telemetry link.

The biggest problem is that whilst we can take all the precautions available, as soon as something goes wrong the RC model in question is a "drone" when it hits the news-stands :(
Log In to reply
Holski77 on March 15, 2014
Why don't you just put landing hear on it? No more hand launching.

Log In to reply
Red20RC on March 16, 2014
If I was building a scale model of a full-size UAV then that would certainly be the answer. The problem is that 99% of "missions" flown by small UAVs take place nowhere near anything resembling a runway. In this case landing gear is not only redundant, it's an added weight/source of drag that we can't afford in a design.
Log In to reply
Santoso on March 16, 2014
Have you ever seen insitu scan eagle..? Find it on youtube.. Maybe can help you..
Log In to reply
Red20RC on March 16, 2014
I have, and I believe it is in use here in Australia.
Interesting design with a constant chord swept wing. I'm guessing though that you are talking about the rail launcher that they use with it?
In fact, you are the second person that has mentioned the Scan Eagle to me as an interesting concept.
Log In to reply
Santoso on March 16, 2014
Yes.. Take off and return back process...
And also the architecture configuration design..
Log In to reply
rick_harriss on November 4, 2014
The recovery system for the insitu scan eagle is something else! In ww2 they cut aircraft in half with a suspended cable!
Log In to reply
jollyboy1 on March 15, 2014
The problem with a pusher configuraton is that there is no propwash over the wing and control surfaces. My friend and I had developed a 15cm flying wing small UAV and this was one of things we learnt first. Without the propwash the aircraft dropped like a stone.

Also, most of the wing is in 3D flow and not 2D flow as compared to a larger model. This means that there is greater drag, but these wings if designed correctly can achieve much greater angles of attack than their larger counterparts.

The most difficult aspect of the smaller UAVs is that even the weight of the autopilot becomes a big issue. We flew our 15cm aircraft without an autopilot and stability system, and very often we would have orientation issues. This was esspecially problematic when we flew for endurance where I had to navigate that dot in the sky for 20 minutes.
Log In to reply
Red20RC on March 16, 2014
15cm?! Crikey! That's not a small UAV, it's a micro, nay - NANO UAV! I would like to have seen that :-) I tend to work at around 1000 - 1200mm for an average wingspan so I'm certainly not in that class!
No wonder the autopilot was weighing it down! That said, I see HK have just release a "HKPilot Mega Mini", which seems to be an 8 gram APM2.5 clone - might be worth another look ;-)

The issue of propwash certainly lends weight back to the Skywalker and Bixler style platforms as at least the tail feathers are always in moving air so you can retain some control at slow speed (although this doesn't help with the lift situation).
In fact, the issue of slow-speed handling is the reason I have recently modified my wing designs to make the outer section of the ailerons much wider than the inboard sections. This gives them better authority at slow speed.

I had some first hand experience with airflow on wings today with a KFm-6 flying wing that I am testing - but that's for a different article...
Log In to reply
jollyswagman on March 16, 2014
I wonder why more people don't use twin tractor setups - al la the FT Kraken.
For the weight of an extra ESC and marginal penalty of two small motors versus one big one you get:
props in clean air,
airflow over the wings,
forward visibility,
no torque roll,
potential for differential thrust to help in turns (e.g. rescuing a bad takeoff!) important for flying wings with limited yaw control,
possibilty of stable vertical takeoff,
less noise
Log In to reply
Red20RC on March 17, 2014
That's a really good question!
I would guess that most people still see twins as having weight and power demands in excess of a single motor. Maybe even twins are still seen as troublesome when setting up. Both of these as a throwback to our old i.c. motor days.

Of course, this is all nonsense! The Kraken has proven the benefits of two motors in an FPV wing. As for setting up; we're all more than happy to strap 3, 4, 6, even 8 motors to a couple of sticks and a KK2 if they are pointing upwards - so what's the problem!

I did see an old TBS video the other day where they were chasing guys in wingsuits. The wing they were using in that (I guess a modified Zephyr) had two small pushers.

Thanks for bringing this up!
Log In to reply
dz1sfb on April 7, 2014
Red,
Just a few observations. You may already be aware of them.
1. Pusher configurations in a conventional plan-form tend to create a number of problems. Most notably is the heavy motor on the rear requires component placement much further ahead of the intended CG position. This puts puts the mass of the model away from the CG making the model more sluggish to respond to control inputs, especially in pitch.
2. The effect of moving components way forward and a short tail moment creates a lot of lateral area ahead of the CG. This can only be countered by having a lot more vertical fin area behind the CG. Essentially the area behind the CG has to be greater than the area ahead of the CG. Or it may be stated that the Center of Pressure must be behind the Center of Gravity. This is what was giving you fits on coordinated turns and sudden control issues in turns. It is on the margin of yaw stability.
3. Wingtip launches are much safer and more controllable with pusher props. Once you get the hang of it, you will not want to do anything else.

I appreciate you article,
Ken :)
Log In to reply
Red20RC on April 8, 2014
Hi Ken,

Thanks for the comments. Your observations have been very much following my own train of thought over the past month since I wrote this article. Particularly with regards to weight distribution and yaw stabilisation.
I'm playing around a lot more with wings and they have a set of unique problems to overcome. My main problem is that most of my design work suffers from two problems:
1 - I get carried away with aesthetics and forget about function sometimes
2 - I tend to do my designing when I really should be asleep!

I can definitely agree on the wingtip launches. It's either that or overhead with all my wings now. Not sure why I ever stopped really as I remember wingtip launching my electric Zagi 10 years ago - I loved that model :-)
Log In to reply
Damig on June 18, 2014
Hey Red, compiling my parts, plans and tools to build. Whoa! thanks for helping with the steep learning curve.

>>>You said: "2 - I tend to do my designing when I really should be asleep! -"

I did: "Spit take with my coffee whilst catching up on the site!!" ROFLMAO

thanks, you do a really fine job RED. scott
Log In to reply

You need to log-in to comment on articles.


Small UAV Design